| Summary: | 碩士 === 輔仁大學 === 應用統計學研究所 === 86 === ABSTRACT
The purpose of this article is to explore the relationship between the criminal suspects' behavior and the individual variables as well as the personal factors of criminal. Data is the criminal suspects' records for those who are the citizen of Keelung City from 1994 to 1996. The record of the second offense has been exclude. Statistical methods, cross tables, Chi Square Automatic Interaction Detector (CHAID) and loglinear model, were applied to analyze data.
The results of data analysis indicate that the personal variables, gender, age, occupation and education, have the significant influence on criminal behavior. The statistical analysis also shows that the variables of criminal suspects' behavior and the personal variables have significant difference. CHAID's found males and teenagers tend to offend property crimes; males for those who do not have occupations
are tend to offend drug crimes. The data of 1995 and 1996 indicates the interactions between the behavior variables and occupation variable is significant. Some problems for future research and conclusions are list in the last chapter.
KEYWORDS: Contingency analysis, Chi Square Automatic Interaction Detector (CHAID) and loglinear model.
|
|---|
